COVID-19 Growth Rate Prediction
Predictions of COVID-19 Growth Rates Using Bayesian Modeling
These are the countries included in the model:
Model Diagnostics - Trace Plots
The following trace plots help to assess the convergence of the MCMC sampler. You can safely ignore this if not familiar with MCMC.
About This Analysis
This analysis was done by Thomas Wiecki. Interactive visualizations were created by Hamel Husain.
The model that we are building assumes exponential growth. This is definitely wrong because growth would just continue uninterrupted into the future. However, in the early phase of an epidemic it's a reasonable assumption.1
We assume a negative binomial likelihood as we are dealing with count data. A Poisson could also be used but the negative binomial allows us to also model the variance separately to give more flexibility.
The model is also hierarchical, pooling information from individual countries.
This notebook gets up-to-date data from the "2019 Novel Coronavirus COVID-19 (2019-nCoV) Data Repository by Johns Hopkins CSSE" GitHub repository. This code is provided under the BSD-3 License. Link to original notebook.↩